Abstract
Objectives
Scaffolds for bone tissue engineering (BTE) are three-dimensional (3D) porous matrices that provide the necessary sites for cell adhesion and proliferation, where the architecture plays an important role. Ideally, BTE scaffolds should have an interconnected network of both large and small pores to facilitate the infiltration of cells and the diffusion of growth factors and nutrients1. Scaffolds for BTE should also enhance osteogenic differentiation to improve bone regeneration. Graphene oxide (GO) can promote osteogenic differentiation of mesenchymal stem cells (MSCs) because it can provide biophysical cues and adsorb biological factors2. However, due to the lack of fabrication strategies, it remains a challenge to develop GO and GO composite scaffolds with a hierarchical architecture for BTE. In this study, we aim to develop a dual-templating method to control the assembly of GO sheets, in order to design and achieve this architecture.
Experimental Methods
We modified the amphiphilicity of GO by adding different amounts of cetyltrimethylammonium bromide (CTAB), hydroxyapatite (HA), polyacrylic acid (PAA) and elastin from bovine neck ligament. Then we developed GO-CTAB, GO-CTAB-HA, GO-CTAB-PAA and GO-elastin stabilized oil in water emulsions. Afterwards, we froze the emulsions at different temperatures (-190 oC and -20 oC), and upon freeze-drying we produced free-standing scaffolds. We reduced GO-CTAB scaffolds in different conditions to obtain rGO scaffolds. We studied the formation of emulsions and the structural and chemical properties of scaffolds through optical microscopy, scanning electron microscopy (SEM), X-ray photoelectron spectroscopy, attenuated total reflectance Fourier transform infrared, and compressive strength tests. We seeded mouse bone marrow MSCs on 2D GO substrates and 3D GO-based scaffolds. Then, we compared the cell proliferation after 10 days of incubation. We also analyzed the cells on GO-based scaffolds after 7 days of incubation using SEM and confocal microscopy, in order to study cell morphology, attachment and infiltration in the scaffolds.
Results and Discussion
The scaffolds based on GO, GO-HA, GO-GO-PAA, GO-elastin and rGO have interconnected primary pores of 150-300 µm in diameter (Figure 1 a-j), which matches the size of the hexane droplet templates. Freezing at -190 oC results in smaller secondary pores around 10 µm in diameter (Figure 1 a-e) while freezing at -20 oC results in larger secondary pores with diameters around 30-50 µm (Figure 1 f-j). This result shows that the secondary pore formation is controlled by ice nucleation and growth in the aqueous phase of the emulsions at different temperatures. Scaffolds based on GO-HA, GO-PAA and especially GO-elastin show an improved modulus of compression compared to GO-based scaffolds (Figure 1 k). The cell culture using MSCs reveals cell proliferation up to 10 days on 2D GO substrates and 3D GO-based scaffolds (Figure 1 l). Compared to 2D substrates, the large and interconnected pores in 3D scaffolds contribute to the higher cell proliferation result. The cells are well spread on the scaffolds with an elongated shape and filamentous extensions at day 7 (Figure 1m). They infiltrate all the way to the center of GO-based scaffolds through primary pores (Figure 1 m, n). These results indicate that primary pores can facilitate the cell infiltration.
Conclusions
The rapid development and broad application of graphene-based porous materials requires efficient and facile fabrication methods. We developed a novel strategy to control the assembly of GO sheets using emulsions and ice templates to fabricate GO-based scaffolds with an interconnected and hierarchical porous structure. The ability to fabricate rGO and GO composite scaffolds with similar architectures demonstrates the versatility of this strategy. The GO-based scaffolds show biocompatibility and allow cell infiltration. We expect that this strategy can lead to improvements in the fabrication of GO-type scaffolds for BTE.
References
1. Karageorgiou V et al. Biomaterials 26 (27):5474-91. 2005
2. Lee WC et al. Acs Nano 5 (9):7334-41. 2011
Acknowledgements
We acknowledge support from NSERC, the Canada Research Chair Foundation and McGill Engineering Doctoral Award.
Figure 1